1. Field of the Invention
The present invention relates to a technology of locating information and a technology of obtaining information in a service, in which various information is obtained through a network, represented by the World Wide Web (WWW) service of the Internet.
2. Description of the Related Art
The conventional communications technology provides a system of establishing communications by specifying a subject (a person, a communications device, a computer, etc.). Therefore, a user performs the following processes through a network as shown in FIG. 1 to obtain necessary information.                checking (retrieving) necessary information in the network; and        obtaining (transferring) the information from a computer storing the information through a relay device in the network        
In the above mentioned method, the quality of a retrieval result (quality of obtained information, accessibility through a network, etc.) largely depends on the ability of a user. Therefore, for example, necessary information may be inefficiently transferred from a remote place without knowing that it is also located close to a user in a network (that is, stored in a place accessible in transfer or at a high speed), Such inefficiency brings not only a disadvantage to a user but also heavy load to the network and the computer, thereby bringing disadvantages to the entire system.
As a conventional technology developed to solve the above mentioned problem, a retrieval service (a search engine, for example, ‘YAHOO as a service of the Internet’) is used in a retrieving process, and an agent service (a proxy service) is used in a transferring process.
A retrieval service relates to a database system, and retrieves an address in a network of computers containing information related to a keyword assigned by a user, and returns the result. The user determines a computer having necessary information based on a plurality of retrieval results, and requests the computer to transfer the information.
An agent service is also referred to as a cache service. That is, the information previously transferred by a user is cached for a predetermined period, and the cached information can be transferred without issuing a request to the destination computer when the second and subsequent requests to transfer the information are issued. If the user can be provided with an agent service in the neighborhood in the network, the above mentioned inefficiency of transferring the information from a remote computer in the network can be avoided.
However, in the conventional technology described above, a retrieval and a transfer are performed independently of each other, there is an inefficiency which can not be avoided.
For example, although a database of the retrieval service is updated frequently, a computer containing network addresses obtained through the retrieval results may store no information at all, or a computer containing the network addresses may stop or be removed. In these cases, communications are not established. Normally, the configuration and the position of a computer in a network system can be changed. Unless the retrieval service reflects such a change at a practically high speed, a user cannot obtain the information, unnecessary cost arises in transferring information, wasteful traffic is used in the network, and other problems of inefficiency occur.
Even when there are no problems with the update of information, an inefficient process may arise.
For example, assume that results X1 and X2 are obtained by retrieving information X. When an X1 is a remote computer in the network, and an X2 is a computer closer than the X1, the user will select the X2. However, the nearest agent server may store the information about the X1. In this case, it is more efficient to select and transfer the X1.
Thus, it is very difficult to centrally manage the retrieval service including the location the replicas of information in the network.
In summary, the conventional technology has the problem that inefficiency occurs by the difference between a retrieval result and an actual state because a retrieval mechanism is independent of a transfer mechanism.